Abstract. In August 2000, organic selenium in a selenized yeast, (Sel-Plex, Alltech

Abstract Upton, Jr., J. Robert. The effects of selenium supplementation on performance and antioxidant enzyme activity in broiler chickens. (Under the direction of Drs. F. W. Edens and P. R. Ferket) In August 2000, organic selenium in a selenized yeast, (Sel-Plex, Alltech Biotechnology Center, Nicholasville, KY) that contained high levels of selenomethionine was approved by the United States Food and Drug Administration as a source of selenium (Se) supplementation for broiler chickens. The previously reported positive responses to the presence of selenomethionine in Se yeast supplemented feed have increased the interest in use of organic Se in all phases of poultry production. Thus, a series of experiments were conducted to compare the influences of organic and inorganic sources of Se (Se yeast or sodium selenite) on production performance parameters. Body weights, feed conversion ratio (FCR), cut up parts yield as a percentage of carcass weight, drip loss from breast meat, and serum thyroid hormones were measured up to 6 weeks of age. Body weight at 42 days was increased in Se yeast fed broilers as compared to those in the no supplemental Se or sodium selenite treatment groups. The combination of sodium selenite and selenized yeast was no more effective than selenized yeast alone on body weight. FCR were improved by all Se sources, with the selenized yeast and selenized yeast + sodium selenite treatments being superior to sodium selenite only fed birds. On a percentage of carcass weight, yields of viscera, feet and neck were higher in selenized yeast treated birds. Yields of leg and pectoralis major, as percentages of carcass weight, were increased and decreased, respectively, in the selenized yeast treated birds. There was an increase in breast meat drip loss when birds were fed sodium selenite as compared to those fed the selenized yeast or no supplemental Se treatments, suggesting that

prooxidant properties of sodium selenite supplementation may be associated with increased moisture loss from processed breast meat. The serum T 4 levels were higher in birds within the no supplemental Se treatment as compared to those supplemented with sodium selenite or selenized yeast. The ratios between serum T 4 and T 3 indicate that selenized yeast treatment facilitated the extra-thyroidal conversion of T 4 to T 3. The Se status of broilers influenced by dietary Se sources may increase the bird s ability to overcome the adverse effects of reactive oxygen metabolites (ROM). The addition of peroxidized fat to poultry diets produces a state of oxidative stress and can increase the level of ROM that must be reduced in the bird. A second study was conducted to evaluate the positive responses to selenized yeast as influenced by an improved status of the antioxidant enzymes, glutathione peroxidase (GSH-Px) and glutathione reductase (GR). This study evaluated the effects of feeding graded levels of oxidized poultry fat on blood and hepatic GSH-Px and hepatic GR activity in broiler chickens given inorganic or organic forms of dietary Se. Neither fat oxidation nor Se source significantly altered the BW and FCR of broilers, or their activity of hepatic GR. Blood GSH-Px was influenced significantly by both fat and Se source, but the fat X Se source interaction was not significant. There was a Se source effect on the hepatic GSH-Px activity with sodium selenite causing an elevated GSH-Px activity, even in the basal diets with no added oxidized fat. There was no peroxide level effect on GSH-Px in the sodium selenite group, but GSH-Px activity in the Se yeast group did not increase until dietary peroxide level was at the highest inclusion rate.

Because elevated GSH-Px is indicative of oxidative stress, the dietary Se yeast supplementation resulted in better Se status in broilers than in birds fed sodium selenite. Only dietary inclusion of the highest level of oxidized fat was sufficient to impose oxidative stress in the Se-yeast fed birds to induce hepatic GSH-Px activity. These results suggest that the dietary Se supplied in an organic form (selenomethionine) may improve Se status in broilers, leading to greater resistance to oxidative stress than when an inorganic form of Se (sodium selenite) is fed. The results from these studies suggest that selenomethionine from selenized yeast may be a superior form of Se for poultry. The beneficial effects of selenomethionine on production performance and the antioxidant enzyme profile of broiler chickens appear to support this conclusion.

The Effects of Selenium Supplementation on Performance and Antioxidant Enzyme Activity in Broiler Chickens. By John Robert Upton, Jr. A thesis submitted to the Graduate Faculty of North Carolina State University in partial fulfillment of the requirements for the Degree of Master of Science Nutrition and Poultry Science Raleigh 2003 Approved by Chair of Advisory Committee Co-chair of Advisory Committee

ii DEDICATION TO GOD AND MY FAMILY

iii Biography John Robert Upton, Jr. (Robbie) was born June 12, 1978 in Elizabeth City, NC to John R. and Mary H. Upton. Robbie has been involved with agriculture all of his life, getting his start by working on the family farm and continuing further by being active in both 4-H and FFA while growing up. In 1996, he graduated from Camden County High School in Camden, NC. His love for agriculture and science led him to study animal science at NCSU where he received a BS degree in Animal Science and a minor in Agricultural Business Management in May 2000. While at NCSU he completed internships and worked in various disciplines of animal and poultry science. After completing several research internships at the National Institute of Environment and Health Sciences (NIEHS) and NCSU, he decided to further his education with a dual Master s Degree in both Poultry Science and Nutrition under the direction of Drs. Frank Edens and Peter Ferket. Robbie is currently employed at Embrex, Inc. as a Research Specialist in the Research and Development division. Robbie married Lauren Hix on June 29, 2002 and currently resides in Cary, NC.

iv Acknowledgements I wish to extend my sincere gratitude to my graduate committee: Drs. F. W. Edens, P. R. Ferket, and C. R. Parkhurst. This fine group of scientists helped guide me through a very rewarding, yet challenging phase of my life. I am thankful to have worked with each of them; their individual and collective strengths have helped me to fulfill this goal in my life. I express sincere gratitude to Dr. Edens for his diligence to keep me on task, advice, guidance, and encouragement. I am heavily indebted to Dr. Kamel Mahmoud for his technical work on the GSH-Px and GR enzyme assays. I learned so much from him during our short time together. He is a wonderful scientist and person. I could not have completed this degree without his help! I want to thank Sue Mann for being such a wonderful technician and friend. I have enjoyed our endless conversations that helped me with some very difficult decisions in my life. You are a good Christian woman and I am proud to have worked with you and known you. I thank all of my fellow graduate students in both the Poultry and Animal Science Departments for your encouragement, wisdom, and above all friendship. This entire endeavor would have been unbearable without you, your help did not go unnoticed! I extend a special thanks to Daniel Moore, Jody Smith, and Kymberly Gowdy who worked beside me so many days and gave much of their personal time to help me succeed. You guys are the greatest! I am indebted to my parents for their unconditional sacrifices that have allowed me to achieve that which they never had the opportunity to obtain. I would like to thank my

v brother, Ryan, for his love and support. It is a great pleasure to have a supportive family as I have been blessed with. Finally I wish to convey my special gratitude to my wife, Lauren, who stood by me through the happiness and tears. I cannot express the depth of my love and appreciation for your faith and belief in me. Your unconditional love and understanding never faltered and was my inspiration as I faced the many challenges involved with this experience.

vi TABLE OF CONTENTS LIST OF FIGURES.vii LIST OF TABLES..viii LITERATURE REVIEW.1 MANUSCRIPT I: Effect of Selenium Supplementation on Performance and Production Parameters in Broiler Chickens.33 Abstract......33 Introduction....34 Materials and Methods.....37 Animal Welfare..... 37 Animals and Experimental Treatments.....37 Diets... 37 Parameters......38 Statistical Analysis. 38 Results....39 Discussion.. 40 Conclusions....43 References.. 44 MANUSCRIPT II: The Effects of Feeding Oxidized Fat and Selenium Source on Performance, Glutathione Peroxidase, and Glutathione Reductase Activity in Broiler Chickens.54 Abstract.. 54 Introduction....55 Materials and Methods...58 Animal Welfare..58 Animals and Experimental Treatments.. 58 Cytosolic Protein Extraction.. 59 Whole Blood Glutathione Peroxidase and Erythrocyte Glutathione Reductase Analysis 59 Erythrocyte Glutathione Reductase...60 Hepatic Glutathione Peroxidase and Glutathione Reductase Analysis..60 Statistical Analysis.....60 Results....61 Discussion..62 Conclusion. 65 References..66 General Summary..78

vii LIST OF FIGURES Manuscript I Figure 1. Effect of selenium source on drip loss from male broiler breast meat.50 Figure 2. Effect of selenium source on serum thyroxine in broiler chickens..51 Figure 3. Effect of selenium source on serum triiodothyronine in broiler chickens...52 Figure 4. Effect of selenium source on serum T 4 /T 3 ratios in broiler chickens...53 Manuscript II Figure 1. Oxidation of poultry fat incorporated in experimental diets...77

viii LIST OF TABLES Manuscript I Table 1. Influence of selenium source on body weight, feed conversion, mortality, and feather weight.47 Table 2. Influence of selenium source on percent yield of parts based on carcass weight 48 Table 3. Influence of selenium source on percent yield of parts based on carcass weight 49 Manuscript II Table 1. Diet composition 72 Table 2. Influence of selenium source and oxidized fat on broiler chicken performance. Data are expressed as means and SE. 73 Table 3. Influence of selenium source and oxidized fat on erythrocyte GSH-Px. Data are expressed as nmol/mg Hb 74 Table 4. Influence of selenium source and oxidized fat on hepatic GSH-Px. Data are expressed as nmol/mg of total protein...75 Table 5. Influence of selenium source and oxidized fat on hepatic GR. Data are expressed as mu/mg of total protein.76

1 Literature Review Introduction Trace elements are essential for the maintenance of health, growth, and a myriad of biochemical-physiological functions (Scott et al., 1982). Selenium (Se) is an essential trace element that was discovered in 1818 by Berzelius in Sweden. No biological significance was associated with Se until it was identified as a toxic agent involved with alkali disease in the Dakota and Wyoming territories of the United States in 1856 (Franke, 1934). It was considered a dangerous element until 1957 when Schwarz and Foltz reported that Se was an essential trace element. Nutritionists and scientists then initiated extensive studies to discover the metabolic function of the element and document the consequences of its deficiency in human and animal diets. However, it was not until 1974 that Se was added as a supplement to poultry and animal diets. Se and sulfur share similar chemical structure and it has long been postulated that they would follow similar pathways of metabolism. The discovery that plants and bacteria metabolize Se to selenocysteine and selenomethionine (SeMet) strengthened this belief (Burnell and Whatley, 1977). Organic Se, found primarily in the form of SeMet is metabolized in the same manner as methionine (Wolfram, 1999). SeMet is readily utilized as a substrate by enzymes that use methionine, and SeMet may be more available than pure methionine (Markham et al., 1980). SeMet is actively transported through intestinal membranes during absorption and actively accumulated in the liver and muscle. Although inorganic Se is absorbed as a mineral, little is retained in the tissue and much of the Se is excreted via the feces in ruminants and urine in the non-ruminant. Very little of the inorganic Se is incorporated into body proteins (Wolfram, 1999).

2 In nature, Se exists in either organic or inorganic forms. Elemental Se can be reduced to the Se -2 oxidation state (selenide) or oxidized to the Se +4 (S0-2 3, selenite) or Se +6 (SO -2 4, selenate). Therefore inorganic Se can be found in the different inorganic forms of selenite, selenate, and selenide. In contrast, Se in feed ingredients is an integral part of the amino acids methionine and cysteine and exists in the Se -2 oxidation state. Therefore, in nature, most animals primarily receive organic Se in the form of SeMet (Combs and Combs, 1986). Plants absorb Se from the soil in the form of selenite or selenate and synthesize selenoaminoacids with SeMet representing about 50% of the Se in cereal grains (Olson and Palmer, 1976). Absorption and Metabolism. Se absorption in the intestine is dependent upon its chemical form. McConnell and Cho (1965) examined everted intestinal sacs of hamsters and found that L-selenomethionine is transported against a gradient from the mucosal to serosal side of the intestine. The transport of L-selenomethionine was inhibited by the corresponding sulfur analogue, L-methionine. Sulfite and cystine, respectively, did not inhibit the transport of selenite and selenocystine. Spencer and Blau (1962) used 35 S- methionine and 75 Se-SeMet to show accumulation of both on the serosal side of everted hamster intestinal sacs. Chromatography was utilized to show only one gamma-emitting peak, suggesting that 75 Se-SeMet was not degraded in transport. Anundi et al. (1984) used in situ intestinal loops, everted gut sacs, and isolated intestinal epithelial cells of rats to demonstrate that intestinal cells concentrate Se from selenite and that reduced glutathione (GSH) in the gut lumen plays a role in selenite absorption. When animals were pre-treated with diethylmaleate, a GSH depleting agent, and

3 when the agent was added to the incubation medium for intestinal cells, the rate of 75 Seselenite disappearance from the lumenal fluid of isolated loops reduced the accumulation of 75 Se from selenite in the cells. Likewise, the addition of GSH enhanced the transfer. In vivo experiments by Reasbeck et al. (1981) using dogs with triple lumen gut perfusion showed that the amount of SeMet absorbed was almost twice that of selenocystine and approximately 4 times that of selenite during a 2-hour test period. This report agrees with the hypothesis that amino acid bound Se is absorbed through specific amino acid active transport mechanisms. Wolfram et al. (1985) investigated the intestinal absorption of selenate and selenite using in vivo perfusion techniques in rats. Different segments of the intestine were perfused, and the concentration dependence of ileal selenate absorption was measured. They found that selenate absorption was greatest in the ileum, where Se was absorbed by a carrier-mediated process. This absorption mechanism was not inhibited by selenite, but it can be inhibited by sulfate. Thompson and Stewart (1975a) found absorption of Se from 75 Se-selenite to average 91-93% and that from 75 Se-SeMet to be 95-97%. In a later report, Thompson and Stewart (1975b) found that the average intestinal absorption of an oral dose of 75 Se-SeMet was 86% and that of 75 Se-selenocystine was 81%. The Se status of an animal appears to have little effect on the intestinal absorption of selenite (Na 2 SeO 3 ). Brown et al. (1972) fed rats varying levels of Se as Na 2 SeO 3 (0, 0.5, 4.0 PPM in a low-se Torula yeast diet) for 33 days before administering Na 75 2 SeO 3, either by stomach tube or intraperitoneal injection. Absorption through the dosage range was 95-100% of the dose. These data suggests the absence of a regulatory mechanism for absorption

4 of selenite because absorption was not different even though a deficient level of Se was supplied for 33 days before administering the varying levels of Se. Although Brown et al. (1972) reported that Se status of the animal might not affect Se absorption, Se status with respect to other elements may. Mykkanen and Humaloja (1984) reported that feeding 1000 PPM of Lead (Pb) to chickens for 3 weeks before absorption measurements of 75 Se-selenite resulted in a reduction in selenite transfer into the body by increasing retention in intestinal tissue. They also showed that increasing the dietary Se alleviated the inhibition. The authors hypothesized that long- term Pb exposure increased the synthesis of proteins that can bind Se, and that these binding sites may become saturated as Se intake is increased. Whanger et al. (1976) demonstrated that no selenite or SeMet is absorbed from the rat stomach. They further reported that selenite and SeMet are absorbed from all segments of the small intestine, with duodenum absorption being slightly higher than from the ileum and jejunum. The specific role of the chick duodenum in the absorption of Se (selenate or selenite) was shown by Apsite et al. (1993). Selenite is passively absorbed in the intestine with the highest concentrations found in duodenum, liver, and kidneys (Apsite et al., 1994). Pesti and Combs (1976) measured accumulation of 75 Se in segments of chick intestine administered Na 75 2 SeO 3 and found a decrease in accumulated radioactivity from the anterior to the posterior of the gut. The importance of the upper intestine in Se absorption was verified by challenging chicks with intestinal coccidia known to infect the chick intestine at specific sites. The coccidial species that infected the duodenum and upper ileum increased the incidence and severity of exudative diathesis (ED), thus lowering Se status and reducing protein binding of 75 Se in the duodenum. Species that infect the posterior ileum had no effect

5 on manifestation of ED or on absorption of Se. Absorption of 75 Se from selenite, selenate, and SeMet was determined in ligated loops from duodenum, jejunum, and ileum of Sedeficient rats or rats fed selenite-supplemented diets. Se deficiency had no effect on absorption of any selenocompound in any intestinal segment. SeMet was absorbed from all segments. In comparison, selenate and selenite were most efficiently absorbed from the ileum (Vendeland et al., 1992). Absorption studies with ligated duodenal loops or oral doses indicated that high vitamin A intake (Combs, 1976) or dietary ascorbic acid (Combs and Pesti, 1976) promoted the enteric absorption of Se. Combs (1978) reported that Ethoxyquin was effective in alleviating ED when fed separately from Se. It was also effective in promoting Se utilization by increasing the plasma concentrations of the Se-containing enzyme, glutathione peroxidase (GSH-Px)(Combs, 1978). Bioavailability. There has been much discussion concerning the bioavailability of Se. Much of this discussion revolves around the Se source. Selenite and SeMet differ in the mechanism in which they are metabolized to selenocysteine for incorporation into GSH-Px. Additionally, SeMet can be incorporated directly into body proteins and stored as such, and therefore could, inflate the bioavailability estimates based on body retention (Henry et al., 1995). Cantor and Scott (1974) examined biological availability of Se for the prevention of ED. In that study, biological availability was taken as 100% for sodium selenite. It ranged between 74% for sodium selenate to 7% for elemental Se. The biological availability of Se in feedstuffs was from 210% for lucerne meal to 60% for soybean meal and in animal feedstuffs from 25% for herring meal to 8.5% for fish solubles. Conversely, SeMet was four times

6 more effective than either selenite or selenocysteine with respect to prevention of pancreatic degeneration and increasing the Se concentration and pancreas weight (Cantor et al., 1975). Cantor and Scott (1974) observed that protection against ED was closely related to plasma GSH-Px. SeMet significantly increased egg production in the 3rd and 4th weeks and hatchability was significantly increased in the 2 nd, 3 rd, 4 th, and 5 th weeks by either supplement. After another three months of the basal diet fed alone or with selenite, egg production was 56 and 77% on the basal and selenite-supplemented diets, respectively, while hatchability was 10 and 90%, respectively. In prevention of ED, Se in most feeds of plant origin was highly available, from 60 to 90%, but was less than 25% available in animal products (Cantor et al, 1975). It was concluded that relative to Na 2 Se 3 O the retention of Se in fish meal or solubles was 43%, and relative to SeMet, 31% (Miller et al, 1972). Hassan et al. (1993) observed that the bioavailability of Se in wheat and fishmeal in comparison to selenite for increasing the activity of GSH-Px was 78% and 58% respectively. Se in fishmeal was poorly available and did not prevent Se deficiency in chickens (Martello and Latshaw, 1982). Ikumo and Yoshida (1981) measured availability of Se in soybean meal, lucerne, fishmeal and SeMet, based on plasma GSH-Px and found them to be 33, 85, 82, and 92% respectively. The results of Whitacre and Latshaw (1982) clearly showed that commercially prepared fishmeal significantly decreased Se utilization and bioavailability. Availability of Se in varying feed sources was evaluated by measuring the restoration of blood serum GSH-Px in Se-depleted chicks. The availability of the Se in capelin fishmeal was 48%, mackerel fishmeal 34%, soybean meal 17.5%, corn gluten meal 25.7%, and SeMet 78.3% (Gabrielsen and Opstvedt, 1980). However, it is important to note that there is a

7 problem when using GSH-Px as a marker because only 0.01 PPM of Se can maximize the enzyme s activity. In a study that provided Se as selenite, SeCys, or SeMet, Se-deficient chicks were fed a diet containing Se at 10 µg/kg of diet. Selenite and SeCys had similar effects in promoting weight gain and preventing ED, while SeMet was less effective. In a similar study with Se provided at 60 µg/kg of diet, tissues from the chicks were analyzed for Se content. Tissue content of chicks fed selenite and SeCys were similar. SeMet fed chicks had higher Se concentrations in breast muscle and pancreas, but lower concentrations in kidney, liver, and heart (Osman and Latshaw, 1976) because muscle compromises a large percentage of the body and more Se is probably retained from feeding SeMet than from selenite. Tissue Concentration of Selenium. The concentration of Se in organs and tissues depends on the particular tissue considered, the amount and form of Se in the diet, the length of time the diet is consumed, and the species of animal. In young animals, Se concentration can also depend on the level of dietary Se consumed by the dam. Generally, the greater the animal s Se intake, the greater the concentration in a particular tissue. The relationship is not linear. When sodium selenite is fed, the tissue concentration approaches a plateau as the Se concentration in the diet rises. The effect is not the same when SeMet is the Se source; the Se concentration continues to rise to some threshold beyond that of selenite. Latshaw (1975) reported that Se concentration of chicken liver and muscle was doubled by feeding Se in natural feedstuffs versus feeding the same level of Na 2 SeO 3. The result was not the same when measuring blood Se concentration. Osman and Latshaw (1976) observed a great increase in pancreas Se concentration when chickens were fed SeMet as compared to an

8 equivalent amount of selenite. Cantor et al (1982) reported that SeMet greatly increased Se concentrations in pancreas, muscle, and gizzard, but not in liver when compared to selenite. Generally, tissues ranked by Se concentration follow the order kidney>liver> pancreas>heart>skeletal muscle. This trend is observed in most species. In chickens, the general ranking of Se concentration follows the order feathers>liver>kidney> muscle>plasma (Arnold et al., 1973; Echevarria et al., 1988). The muscle concentration of Se is low, but due to its relatively large mass, muscle contains nearly 40% of total body Se. The liver contains about 30% and all other organs and tissues are comprised of less than 10% each (Behne and Wolters, 1983). Excretion. Se can be excreted from the body by three major routes: the lungs, the urinary tract, and the intestinal tract. The amount and distribution of the Se eliminated depends on many factors: Se intake, form of Se, composition of the diet, as well as the physiological status of the animal. Burk et al. (1972) investigated Se excretion in rats using an injected tracer dose of H 75 2 SeO 3. Urinary excretion of Se was directly proportional to the dietary Se level, ranging from 6% with the low Se-supplemented diet to 67% with the high Se-supplemented diet. Fecal excretion of Se remained nearly constant at approximately 10% of the dose as the Se concentration in the diet varied. Respiratory elimination of Se was marginal and constant over the range of dietary Se levels. The authors concluded that the rats adjusted to variations in Se intake by altering the excretion of Se in the urine. The effect of Se form on excretion in rats has been reported by many authors (Thomson et al., 1975a, Thomson et al., 1975b, Wright et al., 1966). Thomson et al. (1975b) gave oral doses of 75 Se to rats as selenite, SeMet, or SeCys and measured urinary excretion of

9 Se. The authors reported excretion during the first week as 4-5% of the dose for SeMet and more than twice that amount for selenite and Se-cystine. Nahapetian et al. (1983) examined oral doses of each of the three forms of Se in the rat. They found that at a dose of 16 µg Se/kg body weight, urinary excretion of Se was similar for selenite and Se-Cystine, but significantly lower for SeMet. When the rats were given a dose of 1.5 mg Se/kg body weight, the urinary excretion of Se was reduced for selenite treated rats to the same percentage as that for SeMet treated rats. However, the SeCys excretion of Se remained unchanged. The decreased urinary excretion of Se from selenite was probably compensated by an increase in respiratory excretion. Urinary excretion is a primary mechanism to decrease Se retention and maintain Se homeostasis. However, there are reports indicating Se is excreted through the feces. Thompson and Robinson (1986) reported that total recovery of Se in urine and feces were similar for both selenate Se and selenite Se. Wright and Bell (1966) examined oral doses of Se 75 in both sheep and swine. In pigs, they found that 22 % of the total oral dose of Se was recovered. Fecal excretion accounted for 15% and 7 % of the recovery as urinary excretion. In sheep, they reported 70 % of the Se 75 was recovered. The majority of the oral Se 75 was recovered in the feces (66%) as were only 4% was recovered in the urine. Ganther et al. (1966) reported manipulation in the amount of volatilization of a dose of Na 75 2 SeO 3 by varying the diet composition of rats. Rats fed the commercially prepared stock diet treatment volatilized about 30% of the dose within 10 hours, where rats consuming a casein-based purified diet exhaled approximately 10%. The authors postulated that with the addition of methionine, increased volatilization occurred due to greater availability of methyl groups for the formation of dimethylselenide.

10 Selenium Deficiency. There are a plethora of diseases associated with Se deficiency in chickens and other animals. The only clearly defined Se deficiency syndrome not related to any other deficiency is nutritional pancreatic atrophy (Combs, 1994). Only under high dietary concentrations of vitamin E (15-20 fold) or the presence of other antioxidants can the chick pancreas be protected in the absence of Se (Whitacre et al., 1987). Other Se deficiencies seem to be exacerbated by vitamin E deficiencies including nutritional encephalomalacia (Century and Hurwitt, 1964; Combs and Hady, 1991) and ED (Noguchi et al., 1973; Barthlomew et al., 1998). Therefore, all three of these Se deficiency diseases are related in some form to dietary antioxidant status. Se deficiency is also associated with impaired development of immunocompetence, and reproductive failure in breeding chickens (Combs and Combs, 1984). Marsh et al. (1981) demonstrated that Se was required for normal development of immunocompetence in the chick by feeding diets deficient in either Se or Vitamin E for the first 2 weeks after hatching. The chicks exhibited impaired humoral responses to sheep red blood cells. However, Se and vitamin E appeared to be mutually replaceable for this function by 3 weeks of age. Combs and Scott (1977) showed hatchability of eggs was reduced by a low-se diet and was further depressed by peroxidized fat. The hatchability was restored to normal by supplementation of Se and Vitamin E. Cantor et al. (1978) showed that Se is required in breeder turkey diets for optimum hatchability and viability of offspring. Egg production and fertility were maintained at 77% and 92%, respectively, by diets supplemented with Se and were reduced to 56% and nearly 0%, respectively, when fed the basal diet that was low in Se (Cantor and Scott, 1974).

11 Selenium Toxicity. Se has been shown to be toxic to poultry when used in high doses. It usually is only seen when the dose exceeds at least 10-fold the physiological requirement. However, Arnold et al. (1972; 1973) reported that the effects of an accidental addition of as much as 80 times the requirement could be quickly rectified. Although there are conflicting reports, Se doses at lower than 3-5 mg/kg of feed are usually not involved with toxicity. Moksnes (1983) fed white leghorn chickens a basal diet containing Se at 0.30 mg/kg. The diets were supplemented with 0, 0.1, 0.5, 1.0, 3.0, and 6.0 mg/kg of Se in the form of SeMet for 18 weeks and no toxic effects observed even at the highest intake of Se (Moksnes, 1983). Kaantee et al., (1982) reported no adverse effect on parent birds or on hatching of eggs when 40 Leghorn hens and five cockerels were fed diets containing 0.14 to 0.85 mg/kg of Se. When Se as sodium selenate was supplemented in feed from 0.1 to 9 mg/kg, hatchability of fertile eggs was significantly decreased by 5 mg/kg Se and higher (Kaantee et al., 1982). Egg weights were decreased by 7 mg/kg and higher and egg production was decreased only by 9 mg/kg (Ort and Latshaw, 1978). Todorovic et al. (1999) fed day old chickens basal diets supplemented with 0, 2, 10, 20, and 30 mg Se/kg as sodium selenite for six weeks. The lowest levels at which dietary Se caused reduction in daily gain was 5 mg/kg. Diets supplemented with 10, 15, and 20 mg Se/kg produced 24.5, 62.7, and 96.6% reductions in daily gain, respectively. Mortality was reported as 26.7, 60, and 80% when diets were fed with 15, 20, and 30 mg Se/kg respectively (Todorovic et al. (1999). Salyi et al. (1993) reported the oral LD 50 of sodium selenite to be 9.7 mg/kg of body mass for chickens. In another experiment the LD 50 for chickens was reported as 24.6 mg Se/kg of body weight (Tishkov and Voitov, 1989). The authors also reported species-

12 specific differences in susceptibility to Se toxicity indicating the LD 50 for turkey poults and ducks to be 13.5 and 64 mg/kg of body weight respectively. In turkey poults the minimum toxic dose rate of sodium selenite was 0.9 mg/kg body weight, as with broiler chicks it was found to be 1.7 mg/kg and 9.4 mg/kg in ducks (Tishkov and Voitov, 1989). Akulov et al (1972) found the LD 50 to be 33.4 mg/kg body weight and the maximum tolerated dose was 15 mg/kg body weight in laying hens. Se toxicity produces characteristic signs such as growth depression, reduced egg production, anemia, and stiffness of the tibiotarsal joints (Soffietti et al., 1993). Qi et al. (1992) fed chicks toxic doses of organic Se (8-13 PPM in Se-enriched corn) and found the pathological changes were characterized by local necrosis in the liver, myocardial degeneration and convoluted tubule necrosis in kidneys. When Se poisoning was experimentally induced by Salyi et al (1993), dypsnea, diarrhea, and weakness were observed within a short time. It is important to note that the molecular mechanisms of Se toxicity are not well defined. However, considerable evidence has accumulated, suggesting that oxidative stress may be the main molecular mechanism of selenosis (Raisbeck, 2000). A reaction between selenite and GSH with production of free radicals could explain the pro-oxidant properties of selenite. Raisbeck (2000) reported that the primary targets of Se toxicity in food animals are the cardiovascular, gastrointestinal, and hematopoetic systems. Selenoproteins. The physiological roles of Se began with groundbreaking work by Rotruck et al. (1973) that identified Se as a stoichiometric component of glutathione peroxidase. Soon thereafter in the mid-1980 s, more selenoproteins were discovered and Se biochemistry began to broaden.

13 Glutathione Peroxidases. The first selenoprotein discovered was glutathione peroxidase (GSH). Ursini et al (1995) described four structurally and genetically different forms of Se-containing GSH-Px that exist in different tissues or parts of the cell. The most abundant selenoprotein in mammalian tissues is the cytosolic GSH-Px (cgsh-px). This classical form of GSH-Px efficiently metabolizes hydrogen peroxide as well as unesterified fatty acid hydroperoxides in conjunction with GSH, the pentose phosphate cycle, and glutathione reductase (GR) (Wolffram, 1999). cgsh-px is a tetrameric protein with four identical subunits, each containing one Se atom (Sunde, 1993). The second cytosolic GSH- Px was identified in the gastrointestinal tract cells and is referred to as GSH-Px-GI. The enzymatic properties of this form are practically the same as those of cytosolic GSH-Px. The physical properties of the two are also similar, with the activity of both depending on Se supply (Chu et al., 1993). GSH-Px-GI activity was reported in both the villus and crypt regions of rat mucosal epithelium, and its activity nearly equaled that of the classical GSH- Px throughout the small intestine and colorectal segments (Esworthy et al., 1998). The third form, which is extracellular, is present in the plasma and denoted as pgsh-px (Daniels, 1996). Phospholipid hydroperoxide GSH-Px (GSH-Px-PH), a membrane-bound form of GSH-Px, plays an important role in the interaction of vitamin E and Se. It has been postulated that GSH-Px-PH along with vitamin E acts as a chain-breaking antioxidant to protect phospholipid membranes. GSH-Px-PH may also be involved in regulation of leukotriene biosynthesis. This enzyme is preferentially expressed in endocrine and reproductive tissues and has been shown to play a major role in male reproduction (Kohrle et

14 al., 2000). Ursini et al. (1999) found that in the testis, the GSH-Px-PH acts as a powerful antioxidant in the developing spermatids and spermatozoa. GSH-Px-PH is distinguished from classical GSH-Px as it is active in monomeric form and has a different amino acid composition (Sunde, 1993). Thioredoxin reductase. Stabilizing disulfides are abundant on cell surfaces and extracellular proteins due to the oxidizing nature of the environment. The inside of the cell is maintained in a very reduced state where intracellular proteins contain many free sulfydryl groups (-SH) and very few disulfides. Thioredoxin is the major disulfide reductase that maintains the proteins in a reduced state. Thioredoxin and GSH work together to keep intracellular redox potential low and maintains free SH groups (Arner and Holmgren, 2000). Thioredoxin is able to transfer reducing power to cellular proteins by the enzyme thioredoxin reductase. Thioredoxin reductase (TR) uses electrons from NADPH to reduce oxidized thioredoxin. All mammalian TRs are Se-dependent flavoproteins (Tamura and Stadtman, 1996). The physiological function of the thioredoxin system is being studied extensively. Mustacich and Powis (2000) reported that thioredoxins are high capacity electron donors for reductive enzymes including ribonucleotide reductase, thioredoxin peroxidase, and through thiol/disulfide exchange reduce key cysteine residues in transcription factors that increase binding to DNA and alters gene transcription. Thioredoxins also function as cell growth factors and have the ability to inhibit apoptosis. Arner and Holmgren (2000) report that TR reduces hydroperoxides, ascorbate and selenite. It has been suggested that ascorbate may be the biochemical link between vitamin E and Se because it has been shown to recycle tocopheroxyl to tocopherol in vitro (Burk and Hill, 1999).

15 Selenoprotein P. Selenoprotein P, along with plasma/extracellular GSH-Px, is the only other known extracellular selenoprotein. Selenoprotein P is unique among the selenoproteins because it contains multiple selenocysteine residues. Selenoprotein P has been isolated from various tissues of the mouse, rat, human, and bovine (Burk and Hill, 1999). Selenoprotein P constitutes a large portion of the Se levels in plasma. It has been shown that more than 60% of plasma Se was found in selenoprotein P in rats (Burk and Hill, 1999) while only 50% in humans (Xia et al., 1992). A specific biochemical function has not been described for selenoprotein P, but both extracellular antioxidant and Se transport roles have been suggested. Burk et al. (1997) reported the association of selenoprotein P with endothelial cells in liver, kidney, and brain of rats. Selenoprotein P binds heparin proteoglycans on cells and in the interstitial matrix since it is secreted into plasma by the liver and into interstitial spaces by cells in virtually all tissues. Burk and Hill (1993) suggest that selenoprotein P functions to protect endothelial cells against oxidants. Selenoprotein P is responsive to dietary Se and is reduced to 5-10% of control levels in rats fed deficient diets. Selenoprotein P has a higher priority for synthesis and declines less rapidly when compared to GSH-Px when animals are fed diets that are limiting in Se (Yang et al., 1989; Burk and Hill, 1993). Selenoprotein W. Selenoprotein W has been isolated from the mouse, rat, monkey, human, and sheep (Burk and Hill, 1993). Although the physiological function is not understood, it is known to be located in the cell cytosol and the tissue occurrence appears to

16 be species dependent (Yeh, et al., 1997a). The highest concentration of selenoprotein W is found in the heart and muscle of sheep. Thus, its relationship to the etiology of white muscle in sheep has been studied. Yeh et al. (1997a) studied sheep fed diets that were deficient in Se or supplemented with selenite (3mg/kg) and investigated selenoprotein W content. All tissue levels of selenoprotein W were sensitive to dietary Se content except the brain. The sheep fed deficient diets had brain tissue levels of selenoprotein W at the same level of those fed the supplemented diets. However, there was a 53% reduction in total brain Se content and a 30% reduction in GSH-Px activity. These results indicated that the regulation of selenoprotein W synthesis in the brain differed from other tissues. In another report by the same authors, it was shown that rat tissues differed in rates of increase in selenoprotein W content in response to graded increases in dietary Se. The report also indicated that there was little correlation with tissue GSH-Px activity or tissue Se content (Yeh et al. 1997b). Iodothyronine Deiodinases. Iodothyronine Deiodinases (ID) are the second largest group of selenoproteins. The three deiodinases (Type I, II, and III ID) control the local availability and concentration of the active thyroid hormone, 3,3,5-triiodothyronine (T 3 ). These enzymes catalyze the conversion of thyroxin (T 4 ) to T 3 (Type I and II ID) or the deiodination of T 4 and T 3 to non-active metabolites (Type III ID). These three isoenzymes are encoded by different genes and have tissue and development-specific patterns of expression and regulation (Kohrle et al., 2000). In general, ID is ranked higher in priority for available Se supply than is cytosolic GSH-Px and was similar in ranking to that for GSH-Px- PH and selenoprotein P (Kohrle, 2000). Because thyroid hormone controls growth, development, differentiation and many metabolic reactions, Se is believed to be involved in regulation of those functions as well.

17 In Se-deficient animals, activity of ID I is low. This results in plasma T 4 increase while T 3 is decreased. A role for Se in type I deiodinase and thyroid hormone metabolism has many implications. The reduced activity of thyroid hormone explains why Se-deficient animals grow more slowly as effects of this hormone are mainly anabolic (Arthur, 1993). Reduced ID I activity in the pituitary is associated with lower levels of growth hormone in Se-deficient animals (MacPherson, 1994). Thyroid hormone activity is a key factor in animal tolerance to cold stress. A wellknown thyroid hormone function is the heat-producing increase in oxygen consumption of tissues in response to cold temperatures. Involvement of thyroid hormones in feathering has long been reported. The active T 3 is known to be intimately involved in feather development. During periods of feather growth, basal metabolic rate increases to provide energy for feather production and to keep the bird warm. Thyroid hormone levels increase and in response the bird increases heat production. If Se is limiting, T 3 levels might be expected to be lower. As feather cover increases, the basal metabolic rate falls and heat production is diminished because oxidative metabolism decreases (Edens, 2000). Glutathione Reductase. Glutathione Reductase (GR) is a ubiquitous enzyme that catalyses the reduction of oxidized glutathione (GSSG) to GSH in the presence of reduced NADPH. GR is essential for the GSH redox cycle that maintains adequate levels of reduced cellular GSH, the predominant form in the tissues, which serves as an antioxidant reacting with free radicals and organic peroxides (Rall and Lehninger, 1952). It is generally accepted that GR is largely responsible for the maintenance of intracellular glutathione in the reduced form. However, the enzyme also functions intracellulary in the reduction of mixed disulphides of GSH and protein.

18 GR is an important enzyme associated with the hexose monophosphate pathway in erythrocytes (Harris and Kellermeyer, 1970). Research has shown GR to be a flavin enzyme with flavin dinucleotide (FAD) serving as an apparent prosthetic group (Scott et al., 1963; Buzard and Kopko, 1963; Staal et al., 1969; Staal and Veeger, 1969). The Se-dependent GSH-Px system, which is vital for the protection against tissue peroxidation, requires the following nutrients: Se as a component of the GSH-Px, the sulfur amino acids in the form of cysteine in GSH, riboflavin in FAD for GR activity and niacin as a component of reduced NADPH. Both GSH-Px and GR are involved in the cycling of glutathione. GSH-Px oxidizes glutathione while reducing toxic endogenous and exogenous peroxides while GR restores the reduced glutathione status. GR activity may be associated with the requirement for the accumulation of oxidized GSH, which is the product of the reaction catalyzed by GSH-Px. (Pablos et al., 1998). Studies with humans (Beutler, 1969; Bamji, 1969; Tillotson and Baker, 1972) and rats (Glatzle et al., 1968; Srivastava and Beutler, 1970; Buzard and Srivastava, 1970) demonstrated that measurement of erythrocyte GR (EGR) activity would be a useful and sensitive procedure for evaluating the riboflavin status of individual subjects or population groups. The EGR activity is altered in vivo by dietary riboflavin and in vitro by FAD. The degree of in vitro stimulation or EGR activity is dependent on the FAD saturation of the apoprotein, which in turn is dependent on the availability of riboflavin. The action of GR is also critically dependent on an adequate supply of NADPH (Godin and Garnett, 1992). Rutz et al. (1988) reported that whole blood GR was reduced in broilers receiving low riboflavin diets. These authors later reported a significant interaction of Se and riboflavin on

19 erythrocyte GR activity coefficient (Rutz et al., 1989). Both Se and riboflavin decreased the coefficient. Brady et al. (1979) reported that erythrocyte GR activity, but not hepatic and muscle GR activity, was increased by riboflavin supplementation in baby pigs. Parsons et al. (1985) reported the percentage active GR declined rapidly when pigs were placed on a riboflavin-unsupplemented diet and was lower than that of riboflavin-supplemented pigs. There are large differences in the activity of both plasma and erythrocyte activity among species reported in the literature. Unlike erythrocyte, plasma GR is not significantly stimulated by the addition of FAD. In species with high plasma GR activity the determination of whole blood FAD activity coefficients may not be a good index of riboflavin status (Board and Peter, 1976). Immunology. Se deficiency has been reported to decrease both cellular and humoral immune function in man and laboratory animals (Combs and Combs, 1986). The knowledge of specific mechanisms in livestock is less detailed than in laboratory animals although the increase in susceptibility to disease in deficient livestock is well documented (Maas, 1998). Sordillo et al., (1997) reported that Se deficiency is an established risk factor in mastitis incidence and has been correlated with decreased bactericidal activity of neutrophils and the severity of mastitis infection. Organic vs. Inorganic Se For many years it has been recognized that the selenoaminoacids: SeMet, selenocysteine, and selenocystine, are the sources of naturally occurring Se in plant based and meat based feed ingredients (Burk, 1976; Levander 1986; Cai et al., 1995). These selenoaminoacids are bound in protein as SeMet and selenocystine and constitute 50-80 % of total Se in plants and grains (Butler and Peterson, 1967) and in

20 Sel-Plex (Alltech, Inc., Nicholasville, KY), an organic Se-enriched yeast (Kelly and Power, 1995). SeMet cannot be directly synthesized from selenite or selenate by animals (Cummins and Martin, 1967; Sunde, 1990). However, selenocysteine can be found in the body of animals fed inorganic Se as selenite and selenate. The presence of selenocysteine is due to the synthesis of glutathione and other selenoproteins, in which the selenocysteine is incorporated. The Se in selenocysteine is incorporated co-translationally using selenide and serine as precursors. The synthesis of selenocysteine involves a unique process where selenide is phosphorylated under the influence of selenophosphate synthetase to selenophosphate. The selenophosphate is made available to a unique seryl-trna SEC that is recognized by selenocysteine synthetase. The selenocysteine synthetase converts seryltrna SEC to selenocysteyl-trna SEC that allows insertion of selenocysteine into a peptide chain (Amberg et al., 1996). The base triplet UGA that ordinarily functions as a stop codon encodes this process of selenocysteine insertion at its appropriate site in the peptide. The selenocysteine insertion also requires a specific mrna, an elongation factor, GTP, and the selenocysteine insertion sequence that all interact at the ribosome to read the UGA selenocysteine codon (Low and Berry, 1996). SeMet is converted to selenocysteine via the enzyme cystothionase (Esaki et al., 1981). Poultry cannot synthesize cysteine de novo so SeMet is required if there is a conversion of SeMet to selenocysteine (Cummins and Martin, 1967; Esaki et al., 1981). Selenocysteine can be substituted for cysteine in many proteins, but it is not incorporated directly into selenoproteins (Sunde, 1990; Daniels, 1996). To allow selenocysteine to be incorporated into specific selenoproteins, there is a requirement for selenocysteine-β-lyase

21 reaction with free selenocysteine to release selenide in the presence of reducing agents (Sunde, 1990; Burk, 1991). Next, Selenocysteyl-tRNA[Ser] SEC, which recognizes the specific UGA stop codons in the selenoprotein-mrna, inserts the new, co-translationally synthesized selenocysteine into the specific selenoprotein (Burke, 1991). Therefore, organic Se must be converted from its original organic to inorganic form and then back to the organic form to fulfill its biological function (Arthur, 1997). This conversion is crucial with regard to the synthesis of selenoproteins because it has been reported that 30 to 80% of the Se in the body may be selenocysteine (Hawks et al., 1985). In instances where SeMet is supplemented at high levels, it can be demonstrated that 40 to 50% of total body Se as SeMet can be found in muscle (Daniels, 1996). Selenocysteine is the pivotal amino acid in the synthesis of Se-dependent cytosolic glutathione peroxidase (Rotruck et al., 1973), but only about 30 % of the body s Se is incorporated into 30 to 100 other selenoproteins in mammals (Burk and Hill, 1993). Positive Responses to Sel-Plex by Broiler Chickens. Sel-Plex, a selenized yeast product by Alltech, Inc (Nicholasville, KY) was recently approved in the United States as a source of organic Se for broiler diets. Sel-Plex provides a cocktail of Se compounds (Kelly and Power, 1995) but SeMet is the primary form of Se in Sel-Plex. Kelly and Power (1995) reports that the organic Se profile in Sel-Plex is similar to the organic Se profile found in plants and grains. Organic Se in Sel-Plex is readily available and is actively absorbed from the intestine via the Na + -dependent neutral amino acid pathway, where selenite is passively absorbed (Schrauzer, 2000). In numerous studies, Sel-Plex has been proven to be superior to selenite as a nutritional source of Se. Edens et al. (2000) reported the induction of feathering causing sex-